Skip to main content

Full text of "Reduction of ortho-nitro-benzoic acid to anthranilic acid"

See other formats


Reduction of Ortho-Nitro-Benzoic 
Acid to Anthranilic Acid 



E. W. Pfafflin and E. F. Winter 








MAY 31, 1921 



PAUL V, GALViN LIBRARY ^^/7/?/ t fi<:,-^.^^ ^^ ^ 

36 WEST 33RD STREET '^- 

Dean of Cultural Studies 

Digitized by tine Internet Arciiive 

in 2009 witii funding from 

CARLI: Consortium of Academic and Researcii Libraries in Illinois 


Ao knowledgment . 


Theory of the Proced-ure, 

Apparatus . 







Realizing that oiir work has only tonchsd 
on this inportant problem and hoping that it will 
"be solved in the near ±\iture by more competent 
scholars of chemistry, we wish to exj.)ress our 
appreciation for the time and effort given iis by 
rrofessor Harry McCorrnack, head of the Chemical 
Engineering Department, -irmour Institute of 
Technologj'-, Chicago, Illinois. 


iieduction is the name usually given to a 
reaction in v/hich oxygen is withdrav.-n fron, or 
hydrogen added to a conpound. In certain cases 
"both of these processes occur. 

In a preliminary study of reducing agents 
we find that reduction can he accomplished in a 
nunher of ways:- Has cent hydrogen in acid, 
neutral and alkaline solution is the most common, 
although it has not met with great success as a 
reducing agent for organic compoxmds. "^lOdiran and 
sodium amalgam are very efficient and may he applied 
conveniently. Metallic tin and stannous chloride 
are also used extensively. i;inc and zinc dust can 
"be used in either acid or alkaline solutions, and 
at times if necessary in neutral sol-ations. 

it is well loiown that reduction converts IIO 
into EH . The agents mostly employed for the 
reduction of nitro compounds are tin and 
hydrochloric acid. The advantage of these bodies 
is the ease of separating the tin by hydrogen 
sulphide; in the solution there remains only the 
chloride of the amino compounds. The amines with a 

basic ciiaracter are precipitated froE their 
acid solutions by ammonia; amino-phenols , by 
sodium carbonate; anino-earboxylic acids, by 
sodium acetate. 

The reduction oi" the ortho-nitro-bensoic 
acid is the removal of the oxygen in the nitro 
group and the substitution of hydrogen in its 
place to form the amino group. The structural 
formula of the basic reaction is as follows-- 



To accomplish such a reduction reaction, it 
is necessary to use some metal which on addition 
to potassium hydroxide will liberate nascent 
hydrogen. v;e found that various metals were 
capable of such action. 

The second reaction in this v;ork would be 
the removal of the metal from the solution. i'Tora 
the properties of the reduced solution and also 


of the metal, v;e found the sulphides were 
insolu'ble in alkaline solution, and also tliat 
the reduced ortho-nitro-benzoic acid v;as 
soluble even in the presence of the sulpiiide. 
T'he metal sulphide precipitate filtered leaves 
tbe reduced ortho-nitro-benzoic acid in 

'[■he third reaction in this work is to 
obtain some kind of solution in which the 
reduced ortho-nitro-bensoic acid is less soliible 
than that of the material used to precipitate it. 

A salt of copper seem.s to be the most 
adaptable to this purpose for tv/o reasons:- One 
that the reduced ortho-nitro-benzoic acid is less 
soluble in such a solution than is the salt of 
copper, and second that the copper can be removed 
from the solution by the addition of HpS, forming 
an insoluble precipitate of copper sulphide which 
can be filtered off, This should leave the reduced 
ortho-nitro-benzoic acid, or the ortho-amino- 
benzoic acid, or better named Anthranilic acid in a 
solution in v/hich it can be recrystallized. 


Anthranilic acid can be re crystallized from 
vmter solution, and its purity determined "by its 
properties and its melting point. 

Anthranilic acid is solnble in ether, 
alcohol, and v;ater, to the greatest extent in the 
former v/hich is the only solvent fror, v/hich it 
will crystallize in long needles, the other solvents 
yielding a granular product. It crystallizes in 
leaflets v-hich melt at 144-5° G. The aoueous 
solution shovi's a hlue fluorescence and tastes sv/eet. 

Our prohlem. is to find a cheaper and more 
easily conductahle method for the production of 
snythetic indigo from ortho-nitro-toluol. The 
process has already "been carried to ortho-nitro- 
"benzoic acid hy the oxidation of ortho-nitro-toluol. 
Our work in the main is to reduce this ortho-nitro- 
"benzoic acid to ortho-amino-henzoic acid by a simple 
and ine:q)ensive procedure. The anthranilic acid can 
then he easily converted into indigo "blue. 


The reaotions are as follov/s:- 


O'-yMAff^^^'^^l^^'^ CuS^AO/puJUo O/i^ 



At the present time processes for 
synthetic indigo have, "by the application of 
scientific knovvledge and of untiring efforts, 
"been worked out "by German chemists, "by v/hich it 
is possible to produce indigo "blue on a large 
scale and at a small profit. 


The process is carried out, as is indicated 
b7/ the eqtiations. 

(XaJJ^/umaJjiJCj ojuA 



Thjs would not he a satisfactory process 
in this country for the reason that the materials 
used are expensive and the process requires skilled 
lahor and considerahle time. 

The nit ro- compounds are reduced "by nascent 
hydrogen in acid, alkaline and neutral solutions. 
?or the reduction of each nitro group six atoms of 


nascent hydrogen are necessary and the reaction 
may be e:>rpressed as follows :- 

aiOg -f- 6H — ►UIHg -+- 2H„0 

The amine produced comhines with excess of 
acid to form a salt, e.g. XEffl HCl, from which the 
amine is obtained by the addition of an excess of 
caustic alkali. Tin and hydrochloric acid are 
usually employed, the metal being first converted 
into stannous and then into stannic chloride. The 
complete reaction is as follows :- 
2i:ii0^+ 3Sn-M4HGl — ^SfZ.lfflg.HCl) -h3SnCl4+4H 

The hydrochloride of the amine combines with 
the stannic chloride, forming a stanni chloride 
(RMgHGl)SnCl^. Various methods are adopted in order 
to obtain the free amine from the acid mixture. 
In this case it is necessary to precipitate the tin 
from the diliited acid solution by means of hydrogen 
sulphide. The stannous or stannic sulphide is 
separated by filtration, the amine hydrochloride 
subsequently decomposed with alkali, and the free 
amine extracted with ether. 

The objection to this method v;e found was the 
fact that the ortho-nitro-benzoic acid was only 


slightly soluble in acid sol-ution and therefore a 
corresponding lovi? yield of anthranilic acid could 
he obtained. 

Alkaline solution also favors reduction, 
and since the ortho-nitro-henzoic acid is soluble 
in such solutions, it was used. 


Mechanical stirrers were used to keep the 
solution of ortho-nitro-benzoic acid, potassium 
hydroxide, and the metal in constant agitation. 

jj'or the evaporation of the reduced solution 
of ortho-nitro-benzoic acid, to concentrate and 
crystallize the ortho-amino-benzoic acid, a vacuum 
drier was used. This was a steam heated oven 
connected with a centrifugal pump to exhaust the 
air. This pump rxmning under normal conditions v/as 
capable of maintaining a twenty five and one half 
inch vacuum. 

Most of the equipment used in this work was 
standard labratory equipment, such as sand bath, 
drying oven, hydrometers, funnels, beakers and test 



To reduce the ortho-nitro-benzoio acid, it 
was evident that the process was one which required 
the removal of the ozygen and the addition of the 
hydrogen, xieduction with hydrogen can he 
accomplished only when the hydrogen is in the nascent 
state. Molecular hydrogen will not afford reduction. 
We found that nascent hydrogen can also he generated 
in acid, alkaline, or neutral solutions; so our first 
problem was to determine the solubility of the ortho- 
nit ro- "benzoic acid in acids, alkali, water and alcohol. 

The acids used were those which will liberate 
nascent hydrogen on the addition of some metal, 
namely: hydrochloric, sulphuric, acetic, and other 
organic acids, we found that the ortho-nitro-benzoic 
acid is almost insoluble in any of the acids named. 
Mve grams of the ortho-nitro-benzoic acid v/ere added 
to 100 c.c. of n/10 ECl. The solution was agitated 
for a period of one hour. The acid was then diluted 
and evaporated almost to dryness. The residue taken 
up with hot water and the ortho-nitro benzoic acid 
recrystallized. 4.78 grams of the acid were recovered 
or 95.3070. The results with the other acids were 


somewhat unsatisfactory particularly the organic 
acids. We fo-und later that addition compounds of 
the acids were formed. 

The alkali used in the testing of the 
solubility was potassium hydroxide. (Potassium 
hydroxide was used in place of sodium hydroxide, 
"because of the inability of the storeroom to 
supply the latter). The ortho-nitro-henzoic acid 
was readily soluble in the alkali which did not 
necessitate the determination of the degree of 

Water was also tested but close observation 
was sufficient to report that the acid v/as not 
soluble in the water. 

Conclusions from- the tests on solubility 
and the results which have already been obtained 
and recorded in handbooks, we decided that the 
reduction reaction which we wished to make would 
have to be made in alkaline solution. 

oince the potassium hydroxide proved to be 
the best solvent for the ortho-nitro-benzoic acid, 
and also because it was possible to produce nascent 
hydrogen in such a solution, we next proceeded to 


determine whether alkali of different concentration 
varied the soluljility. The result v/as obvious; the 
stronger the alkali the more soluble the acid, 
iiowever, we found later that the concentration of 
the alkali varied the yield of the addition compounds 
formed. The results of this experimenting proved 
that 5° jjaume potassium hydroxide gave the best 
yields of addition compounds of anthranilic acid. 

The next problem was the selection of a 
metal which would produce nascent hydrogen. Zinc 
and aluminum were the metals selected to afford the 
liberation of hydrogen for the reduction. It v/as 
evident at the outset that each of these metals 
v;ould produce the required results, but 
complications which arose later limited the selection. 
The aluminiffii proved unsatisfactory because of two 
reasons: one that the aluminum became coated with 
a film of oxide which prevented trie further 
liberation of hydrogen; and second we found it 
impossible to secure any suitable means of removing 
the aluminum from the solution after the reduction 
had been completed. 


The metal zinc was then 1186(3. in the 
granular state because of its ability to 
liberate the hydrogen and also because an easy 
and efficient method for its removal from 
solution was known. Ciuantative runs proved that 
very little of the zinc entered the solution, 
ciamples of 5 grams of the ortho-nitro-benzoic acid 
was dissolved in 125 o.c. of the 5° Baume potassium 
hydroxide and 2 grams of the granular zinc added. 
Also 2 grams of the zinc was added to 125 c.c. of 
5° i3aume Potassium hydroxide v/ith the absence of 
the ortho-nitro-benzoic acid. The two solutions 
were agitated for two hours after which reduction 
was complete in the one case and in the second all 
the potassium zincate was formed that the reaction 
called for. The solutions were then heated to 
about 85° 0. and allov/ed to absorb HpS under a 
pressure of two inches of water for one hour. The 
precipitate of zinc sulphide was filtered off, dried 
and vreighed as zinc siilphide. rhe results shov^^ed 
that almost equal amounts of zinc entered the two 
solutions. This part of the work was only carried 
on in order to make a rough estimate of the quantity 


of zinc which it would be necessary to remove if 
the method proved one of commercial importance. 
iJo effort was made on the part of the 
experimenters to prevent the zinc from entering 
the solution "because of the ease with which it 
could he removed in the experimental v/ork. It may 
have been necessary to obtain such a procedure, had 
the results of the thesis "been positive. 

The numerical results were as follows: 
In the solution of ortho-nitro-benzoic 66.23% of 
the original zinc was precipitated in the form of 
zinc sulphide from the zincate solution. 

Zn 2K0H ^2 ^^^2 ^^ 
KgZnOg E S ZnS 2K0H. 

62.50% of the zinc was recovered from the second 
sample which did not contain the ortho-nitro-benzoic. 
Whether the ortho-nitro-benzoic acid favored the 
solubility of the zinc, or whether it was an error 
on the part of the experimenters, was not determined 
because the item carried little weight. 

The next problem which confronted us was to 
make the solution from which the zinc had been 
extracted, one in which it would be possible to 


crystallize the acid which we were looking for. 
This proved to he the most difficult prohlem in 
that anthranilio acid has a great affinity for 
nearly all metal salts, rhe proof of this 
conclusion rested wholly in the salt formed v.'ith 
nlokel sulphate, upon the addition of nickel 
sulphate to the solution of anthranilio acid, we 
found that it was impossible to obtain a 
precipitate of nickel sulphide, because of the 
inability of the nickel to ionize in such a 

We also found that it was possible to 
obtain solutions with other acids which would 
render the anthranilio acid less insoluble. 
Great care was exercised in the evaporation of the 
resulting solutions because of the tendancy of the 
organic acids to form addition and substitution com- 
pounds, and tar. iror this reason we found that it 
was necessary to evaporate the solutions under 
reduced pressure and at a low temperature so that the 
product would not go to tar or be burnt. The vacuum 
oven described in the paragraph on apparatus, was 
used for this purpose. 


Among the aoids whioh were used to 
precipitate the anthranilio acid were citric, 
acetic, oxalic, tartaric, and hydrochloric. 
With the addition of the citric acid no 
precipitate was formed, we drew a conclusion 
from this that the citric acid formed an addition 
compound vdth the anthranilic acid which was 
soluble. The addition of oxalic, tartaric, and 
hydrochloric acids gave similar results. Y/hen 
acetic acid was added to the solution, a 
precipitate was obtained. This precipitate was 
washed from the filter paper and dissolved in v/arm 
water, and allowed to evaporate in the vacuum oven. 
The resulting material was of a spongy amorphous 
nature, yellow in color and having a melting point 
of 198.5° Centigrade. An analysis was not necessary 
to prove its composition, however to obtain the 
satisfaction that the product was an addition compound 
of acetic and anthranilic acids or acet-anthranilate , 
the follovYing method was employed: A small portion 
of the material was mixed with an equal proportion of 
crystals of phosphoric acid. It was then diluted 
with water and distilled in a distilling flask. 


condensing the vapor in a water cooled condenser. 
The theory of this operation was that the phosphoric 
acid "being a stronger acid than the acetic v;ould 
replace the acetic, changing the acet-anthranilate 
to phosphoric-anthranilate. This would liberate the 
acetic acid and allow it to he distilled over into 
the receiving flask. This distillate can be tested 
for the presence of acetic acid by some of the 
standard methods. I'he one which was employed by the 
experimenters was a physical property, namely the 

vie found that acetic acid was present in the 
distillate proving the aforesaid statement, that the 
addition of the acetic acid to the solution of 
anthranilic acid forms an addition compound, i^o 
suitable means was at hand to separate the addition 
compound formed so the method was discarded. The 
addition of the other acids v;ould have presented the 
same difficulty had v^^e tried to isolate the soluble 
addition compounds formed. 

After consulting the literature again we 
found that the real problem in the thesis was the 
separation of anthranilic acid from the salts used 


to precipitate it. Because of the small amoiuit of 
material in the literature, this subject was not 
touched upon so the experimentors decided to find a 
metal salt which would not form an addition compound 
and also allow the anthranilio acid to crystallize 
from the solution. 

Anthranilio acid was prepared "by the 
phthalimide method, as described in the introduction. 
The product was tested for its purity and determined 
to he 98. 3^^ pure. The melting point of the product 
was 145° C. The melting point of the acid given in 
the handbooks is 144° C. 

(Hote ) . All of the melting points determined in 
this work and mentioned in this experiment were 
determined as follows:- A small portion of the 
material was ground to a fine pov;der and inserted in 
a capillary tube. The tube was then fastened to a 
thermometer so that the bulb of the thermometer and 
the bottom of the capillary tube were at the same 
level. This was then immersed in two beakers of 
sulphuric acid, one beaker set inside of the other 
to assure an equal temperature of the acid. Heat 


was then applied to the "beakers, and the acid kept 
thoroughly agitated until the material in the 
capillary tuhe melted. This point can readily be 
seen hy the observer and the temperature also noted. 

I'he anthranilic v;as first tested to determine 
from which solvent it would crystallize best. Alcohol, 
ether, and water were used as solvents, we found 
that the acid crystallized readily from ether and 
alcohol, but not so readily from water. Later, 
difficulties arose which prohibited the use of 
alcohol and ether, so the experimenters were forced 
to use water, however we found it very easy to 
reorystallize the anthranilic acid, 

A small quantity of the acid 7/as then dissolved 
in ?/ater and the following salts were added to each 
of 12 small portions of the anthranilic acid. The 
solutions were then heated and allowed to absorb 
HgS under a pressure of two inches of water for a 
period of one hour. 

Copper nitrate was added giving a precipitate 
of copper-anthranilate of blue color. HpS was then 
passed and the precipitate blackened in color. 

22 < 

supposedly that of GuS. The precipitate was 
filtered off and the filtrate evaporated to 
dryness in the vacuiun oven. The residue was a 
yellow crystalline mass trndouhtedly an addition 
compo-und of copper anthranilate. The melting 
point was found to "be 219*^ C. 

Silver nitrate was added giving a "black 
precipitate of silver anthranilate. HpS was then 
allowed to pass through the solution for one hour 
at a pressure of two inches of v/ater. There was 
no vlsahle change in the nature of the precipitate 
"but it was supposed and afterwards proven that the 
silver precipitate was AggS. The filtrate was 
allowed to evaporate in the vacuum oven which left 
a dirty yellow mass, amphorous in structure and 
having a melting point of 223° C. 

Lead- nitrate was next employed, giving a 
white precipitate. Upon the addition of H S the 
white color turned black. The solution was filtered 
and the filtrate evaporated as before. The residue 
this time was brown having a crystalline form and 


melting at 220° C. 

Mercuric nitrate also gave a precipitate 
white in color, which upon the addition of the 
H„S darkened considerahly. This was likewise 
filtered and the filtrate evaporated as before. 
This tiiBe a white crystalline mass resulted which 
v/hen tested for its melting point gave 232° C 
P03TI0N 5 . 

Baritun was the last of the nitrates to he 
tested for its ability to form addition compoxmds 
with the aathranilic acid. For some unknown reason 
no precipitate v/as formed on the addition of the 
barium salt and on the addition of the H2S only 
sulphxjr was precipitated. Results discarded. 

The salts used in portions 1 to 4 inclusive 
gave only addition compounds, while 5 seemed to have 
no affect whatsoever on the acid solution. 
SilviPLE 6. 

Barium chloride v;as next added to the 
anthranilic acid solution. This salt behaved as 
did the barium nitrate but when the HgS was passed 


a precipitate was formed unliice the one formed 
in the case of sample 5. The precipitate was 
filtered off and the filtrate evaporated. The 
result of this salt was a positive one in that the 
residue left in the evaporating dish resembled 
the anthranilic acid v/hich we had started with, 
and v;hich also had a melting point of 14 6° G, 
Believing that for some unJaiov/n reason this salt of 
barium might behave different under different 
conditions, experimenting was not stopped at this 

Because of the inspiring results obtained 
when the chloride was used, and also from the 
amount of information vfcich v;e had found in the 
literature concerning the salt of copper we decided 
to use copper chloride. Copper anthranilate 
precipitated very nicely, and changed color upon 
the addition of the HgS , but to no avail. '.Then the 
precipitate was filtered and the filtrate allov;ed 
to evaporate in the vacuum oven only a resinous 
tarry mass v;as left. 



Another ahloride was used namely the 
merctacic chloride. This was likewise added to 
the solution of anthranilic acid. A white 
precipitate was formed which turned black upon 
the addition of the HgS . This v/as filtered and 
the filtrate evaporated to dryness in the vacuum 
oven. The crystalline roass which remained proved 
to be anthranilic acid having a melting point of 

147.50 C. 


This sample was tested by the addition 
of aluminum sulphate, but was rapidly discarded 
because of the condition of the aliiminum 
precipitate, which rendered it almost unfilterable . 
SAilPLE 10. 

■Jhen nickel sulphate was used we found 
results v/hich were very much out of the ordinary. 
Nickel anthranilate precipitated, but upon the 
addition of HpS no further change took place. 
This was readily observable because the color of 
the nickel anthranilate was green, and upon 
the addition of HpS it did not turn black, v/hich 


is the color of the HiSg. This proved that the 
nickel anthranilate does not ionize as a nickel 
salt and could not "be precipitated "by the use of 

Zinc sulphate was the salt added to this 
sample "but no precipitate v/as formed either on 
the addition of the ZnS04 or on the addition of 
the HgS. Evaporation of the solution resulted in 
a tarry mass. 
SMPLE 1£. 

xiead suhacetate was lastly used in the 
attempt to recrystallize the acid from solutions 
of salts. The result in this case was an 
amorphous yellow mass which melted at approximately 
287° C. 

Samples 6 and 8 were the only ones which 
anthranilic acid could he recovered from. 

Samples 7,9,10, 11, 12. were discarded for 
the various reasons named. 

The results from the anthranilic recrystalliza- 
tion only lend additional problems in this work. 


Conditions may have altered the results 
materially, "but ov/ing to the niunlDer of addition 
compounds formed we deemed it unnecessary to try 
such a thing as a quantitative run. The size of 
the addition compounds or rather the hulk of them 
was sufficient evidence for us that addition 
compounds were the only things we had thus far 

jfTom the results on the escperimenting just 
descrihed, we proceeded from the original basis 
namely the reduced solution of ortho-nitro- 
"benzoic acid, using all of the data obtained from 
the experimenting on the anthranilic. 

A 5 gram sample of ortho-nitro-henzoic acid 
was dissolved in 125 c.c. of 5° Baume potassium 

Case A. The solution was boiled before agitating. 
Case B. The solution was agitated with no boiling, 
£ grams of granular zinc was also added to each 

The solution was decanted, boiled and 
allov/ed to absorb HgS for a period of one hour 


■under a pressure of two inches of water. The 
solutions were heated again if the temperature 
fell "below 80° U. The precipitate which was 
formed was filtered off, which when tested gave 
positive results for zinc sulphide. 

Ihis is where the process tranches, hecause 
of the addition of different salts to favor the 
crystallization of the ortho-amino-henzoic acid. 

The 5 gram sample of the ortho-nitro- 
l)enzolc acid was treated as in procedure 1, case B, 
and procedure 2, In procedure 3, acetic acid was 
employed. Dilute acetic acid v;as added to the 
solution until it showed an acid reaction v;ith 
phenophthalein, when this point was reached a yellow 
precipitate was formed. The precipitate was washed 
from the filter paper and dissolved in a small 
quantity of hot water, evaporated in the vacuum oven 
and re crystallized from water. The resulting residue 
was an amorphous spongy yellow mass similar to that 
ohtained when acetic acid was added to the solution 
of anthranilic acid. The melting point of this 
hody was 201° G. 


SlIdPLES n and m. 

These were also 5 gram samples of the 
ortho-nitro-benzoic acid, treated as in 
procedure 1, case A and B, and procedure 2. Both 
samples, the one which had been boiled and the 
one which was allowed to react without the aid of 
heat, were treated as follows:- copper acetate 
was added, giving the precipitate of copper 
anthranilate, as it did in the case of the 
solution of anthranilic acid. HgS was then passed 
into both of the solutions and the precipitate of 
copper sulphide filtered off. The filtrate was 
evaporated in the vacuum oven under a reduced 
pressure of £3,5 inches of mercury and at a 
temperature of 68° C. Both of the samples gave a 
yellow amorphous mass which was dried and treated 
with carbon bisulphide to remove the sulphur. This 
caused the mass to change color from the original 
yellow to a dirty white, i'he melting point of 
both of these products was about 187° C. 


SAMPLES IV. V, VI, Vll, Vlll. 

These samples were treated as in procedure 
1-B and 2, and the following salts were added: 
IV Sodium acetate. 
"V lead acetate. 

VI Zinc acetate 

VII Lead suhacetate. 

VIII Ammonium acetate. 

aample IV was discarded because of the 
tarry residue which was left upon evaporation. 

Sample V was re crystallized from water 
and the product tested for its melting point. 
210° C. 

Sample VI was discarded because there was 
no precipitate upon the addition of the zinc 

Sample Vll was carried through by the 
regular procedure but the product showed a melting 
point of 200° G. 

Sample Vlll was discarded because of the 
failure to precipitate upon the addition of the 
ammonium acetate. 

The following salts which were employed were 


either used "because positive results had "been 
ohtained from the saithranilic acid experiments 
or he cause of the property of the salt to form 
addition compounds with the anthranilic acid. 
We thought it necessary to use such salts as 
those which had formed addition compounds, to 
prove that the ortho-nitro benzoic acid had teen 
reduced hy the process to which it had "been 

Copper nitrate was used to precipitate the 
anthranilic acid. After the removal of the copper 
hy the addition of the iipS the filtrate was 
evaporated and the residue tested for anthranilic 
acid. The melting point of the product was 215° C. 
which was only 4° from the melting point of the 
product obtained when the copi:)er nitrate was added 
to the solution of anthranilic acid. It was 
extremely hard to get this compound, i'ive such 
samples v;ere run, four of v/hich resulted in only a 
tarry resinous mass. 

Lead nitrate was next used in an effort to 
obtain the pure anthranilic acid, when the nitrate 


was used in experimenting v/ith the anthranilic 
acid, an addition oompoimd was formed whicTi 
melted at 223° C. Five such samples were used 
in this effort "but each of the five times only a 
tarry mass remained in the evaporating dish after 
evaporation, xemperature as low as 60° C. were 
employed to try and prevent the organic compounds 
from turning to tar, hut v;ith no result. 

Barium chloride was one of the salts which 
had permitted the anthranilic acid to recrystallize 
when added to a solution of the acid. Extreme care 
was taken in the evaporation of the filtrate so as 
to prevent, if possihle, the formation of tar. In 
this case the solution Y/as evaporated and left a 
crystalline mass of yellow material. This was re- 
dissolved in a small amount of hot water and 
re crystallized. The product looked as though we had 
found the suitable salt, hxit the melting point of the 
product, instead of heing 144-5° C. as is that of 
anthranilic, was 210° C, 



copper sulphate was added to this sample. 
The solution was carried through the same 
processes as were the others. The residue was a 
hrown crystalline mass melting at 219*^ C. 

Zinc sulphate gave approximately the same 
results as did the copper sulphate. 

Mercuric chloride, one of the salts which 
had also allowed the anthranilic acid to 
re crystallize was used, rhe sample was carried 
through the same procedures as were those in the 
ahove cases. The residue was recrystallized from 
water solution hut to no avail. The melting point 
was found to he 197° G. The melting point when the 
mercuric chloride was added to the solution of 
anthranilic acid was 146° which is somewhat near the 
melting point of anthranilic acid. 

Since all of the ahove methods failed in some 
way or other, the experimenters decided to try some 
of the other methods used in organic chemistry for 


the separation of the metal from the salt, i'hese 
methods did not offer a very wide field due to 
various reasons, with the use of acid to 
precipitate calcium or "barium, only addition 
compounds would he formed and with the second 
possibility that the organic material would he 
reduced to tar. baits of "barium and calcitan were 
used, however, and an effort made to precipitate 
them from the solution hy the passage of carhon 
dioixide under a pressure of 2 inches of water. 
Conditions of temperature were varied and also the 
time of absorption, hut the precipitate formed 
could not he readily filtered. It was evident 
at the outset from our Icnov/ledge of quantative 
chemistry that either of these precipitates would 
he hard to handle. 

From the results of all of the experimenting, 
it follows that the ortho-nitro-henzoic acid was 
reduced in alkaline solution hy nascent hydrogen 
produced by the action of granular zinc on 
potassium hydroxide. We can draw this conclusion 
from the facts that copper nitrate when added to a 


solution of anthranilic acid gave an addition 
compoimd which resemhled and melted fotir degrees 
from the addition compound formed when the copper 
salt was added to the reduced solution of ortho- 
nitro-henzoic acid, .-e also found that no heat 
was necessary for this reduction, however, heat 
was generated to a small extent in the reaction. 

Even though the reduction can "be 
accomplished we are forced to draw the conclusion 
that anthranilic acid cannot be made by this 
process because no suitable means was found that 
will successfully separate the acid from its salt. 
It is also evident that anthranilic acid has a strong 
affinity for most acids and metals, thereby 
rendering it difficult to obtain it in the free state. 

It may be possible hoYi^ever to make indigo blue 
without first obtaining the free anthranilic acid 
by forming the chloracetic addition compound which 
is really the intermediate product between 
anthranilic acid and indigo blue. 



General iTinciples of Organic Chemistry Alexoyeff. 

Praotioal Organic Chemistry. Sudburough and James. 

Organic Chemistry. Perkins and Kipping. 

Organic chemistry. lloyes. 

Organic Chemistry. Kichter. 

Jahreshuoh iiaurent 589. 

Berichte 19 1402 

Aunalen 39 91 

Annalen 39 83 

Beilstein, Kuhlberg, Ihig 163. 138. 

Journal Chemical Society 37 

Hoogenwerf, Van Dorp, liec, trav, chim, Pays - Bas 10 

.. '^\<^^>f:^